Unformatted text preview:

OPTICS AND Optical Instrumentation Optical imaging is the manipulation of light to elucidate the structures of objects History of Optics Optical Instrumentation – light sources Optical Instrumentation – detectors Physical Optics Optical Instrumentation – intermediate opticsA Historical Snap Shot of Optical Study The study of light has a long history dating back to far antiquity. Optical microscope was first invited in the 16 century. However, we will focus on the beginning of 20th century where there are two class of thoughts about the physical properties of light. Wave Nature of Light -- Huygen Particle Nature of Light -- Newton Figures from Wikipedia.History of Optical Studies The advent of quantum mechanics allows us to understand that light has both wave and particular properties Planck – quantization of black body radiation Courtesy of the Clendening History of MedicineLibrary,University of Kansas Medical Center.00500 1000 1500 2000200400600800T = 5500kT = 5000kT = 4500kT = 4000kT = 3500kλ [nm] u(λ ) [kJ/nm] Figure by MIT OCW.History of Optical Studies The advent of quantum mechanics allows us to understand that light has both wave and particular properties Bohr – Resolve wave-particle duality of light Niels Bohr debating quantum theorywith Albert Einstein.Photo byPaul Ehrenfest (source: Wikipedia).n = 4n = 3n = 2n = 1Line SpectrumWavelengthFigure by MIT OCW.A typical biomedical optics experiment Light source Detector Intermediate Optics SpecimenPhysical Principle of High Sensitivity Optical Detectors High sensitivity photodetectors today are mainly based on two physical processes: (1) Photoelectric effect (2) Photovoltic effect One can detect light by other processes such as heating. Power meter for laser light is called a thermopile and is based on heating by light – not very sensitivePhotoelectric Effect First observed by Becquerel in 1839, he observed current in conductive solutions as electrode is exposed to light Theoretically explained by Einstein: An electron knocks out of a material by a photon. It is one of the major evidence in the quantization of light. hn = f + Ek f is the work function characterizing the barrier in the material for electron Ejection. Ek is the kinetic energy of the ejected electron. The kinetic energy depends only on the color (energy) of the photon but not light intensity (number of photons) The number of electrons ejected is proportional to the number of photonsPhotovoltic Effect QM predicts that the electrons in a periodic lattice occupy energy bands that has gaps.Photovoltic Effect II Photovoltic effect: Electron, hole generation in semi-conductor material by lightSignal and Noise in Optical Detection Signal – the amount of light incident upon the detector per unit time n is the number of photons detected per unit time Dt is the data acquisition time < I >= anq / Dt -q is the electron charge= 1.6·10 19 C (1A = 1C/sec) a is a gain factor of the detector Noise – the “disturbance” on the signal level that hinders an accurate measurementSignal-to-Noise Ratio and Noise Equivalent Power Signal: S =< I >2 R SNR: Signal power/Noise power = S/N NEP: Signal power at which SNR = 1Source of Noise in Optical Detectors (1) Optical shot noise (Ns) – inherent noise in counting a finite number of photons per unit time (2) Dark current noise (Nd) – thermally induced “firing” of the detector (3) Johnson noise (NJ) – thermally induced current fluctuation in the load resistor Since the noises are uncorrelated, the different sources of noise add in quadrature N 2 � NS 2 + Nd 2 + NJ 2Optical Shot Noise Photon arrival at detector are statistically independent, “uncorrelated”, events What do we meant by uncorrelated? 1 T /2 Lim �(n(t +t ) - n)(n(t) - n))* =< Dn(t +t )Dn*(t) >= 0 t „ 0 T fi¥ T -T /2 (* denotes complex conjugate) Although the mean number of photons arriving per unit time, l, is constant on average, at each measurement time interval, the number of detected photons can vary. The statistical fluctuation of these un-correlated random events are characterized by Poisson statistics.Poisson Statistics If the mean number of photon detected is n , the probability of observing n photons in time interval t is: n -nP(n | n) = en n! Mean: =1 =5 =10 =20 n n n n 1 M n =�ni Mi Variance: 1 M s n 2 =�(ni - n)2 Mi 2 nn s=Spectrum of Possion Noise I ¥ D ~ I ( f ) = �DI (t)e -i2pftdt where DI(t) = qDf (n(t) -n) -¥ Assume photon number is Poisson distributed Power spectral density: P ~( f ) = RDfDI ~*( f )DI ~( f ) Noise power: N ~( f , Df ) = P ~( f )Df The power spectral density can be evaluated in a slightly round about way by considering the autocorrelation function: ¥ Autocorrelation function: g(t ) = RDf �DI(t +t )DI(t)*dt -¥ Because the event of Poisson process is completely independent of each other g(t ) = Rs I 2d (t )/ DfSpectrum of Poisson Noise II d (t ) is the Dirac-Delta function with the following properties: It has the unit of frequency d (0) =¥; d (t) = 0 for t „ 0 �d (t)dt =1; � f (t)d (t -t )dt = f (t ) From Poisson process: s I 2 = 2aqDf < I > Factor of 2 account for positive and negative frequency bands The autocorelation function of Poisson noise is: g(t ) = 2Raq < I > d (t )Spectrum of Poisson Noise III ¥ Wiener-Khintchine Theorem: P ~( f ) =� g(t )e -i2pft dt -¥Let’s why Wiener-Khintchine theorem is true: ¥ ¥ ¥ g(t )e -i2pft dt = RDf [ DI(t +t )DI(t)dt]e -i2pft dt� � � -¥ -¥ -¥ ¥ ¥ � � -i 2pft= RDf [ DI(t +t )e dt ]DI(t)dt -¥ -¥ ¥ ¥ = RDf [ DI(t ')e -i2pft ' dt ']e+i2pftDI(t)dt� � -¥ -¥ t ' = t +t , dt ' = dt ¥ ¥ = RDf [ �DI(t ')e -i2pft ' dt '][ �DI(t)e+i 2pftdt] -¥ -¥ = RDfD ~ I ( f )DI ~( f )* Fourier transform of the autocorrelation function is the power spectral densitySpectrum of Possion Noise IV ¥ P ~( f ) =�2Raq < I > d (t )e -i2pft df =2Raq < I > -¥ ft Poisson noise has a “white” spectrum Noise in a given spectral band: ~ N( f ,Df ) = 2Raq < I > DfPhoton Shot Noise The origin of the photon shot noise comes from the Poisson statistics of the incoming photons itself The shot noise power is: N~ s ( f , Df ) = 2Raq < I > Df Log(S/N) The signal power is: S =< I >2 R < I > aqn / Dt 2aqnDfSNR = = = = n 2aqDf 2aqDf 2aqDf Used sampling theorem: 1/ Dt = 2Df A detector is consider to be “ideal” if it is dominated by just shot


View Full Document

MIT 20 309 - OPTICS AND Optical Instrumentation

Download OPTICS AND Optical Instrumentation
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view OPTICS AND Optical Instrumentation and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view OPTICS AND Optical Instrumentation 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?